ai engineering agents
5 open-source ai agents from the octomind tap. the command below starts the lead; a specialist starts with its own tag. or run them in the cloud with nothing to install.
octomind run ai starts AI Lead – Build, Evaluate, Secure, Govern, which answers or hands the task to one of the 4 specialists below.
-
Covers LLM and RAG apps, evals, red teaming, and AI governance.
ai
specialists
AI Engineer – LLM & RAG Application Builder
Builds production LLM applications: RAG, agent tool schemas, context engineering, prompt caching, tracing, and MCP integration.
ai:engineerAI Evaluation Engineer – LLM Evals
Designs LLM evaluation suites: golden datasets, RAGAS and G-Eval metrics, statistical significance testing, drift detection, and CI regression gates.
ai:evalsAI Governance Navigator – EU AI Act Compliance
Classifies AI systems under the EU AI Act, maps obligations by tier and role, and builds an AI governance framework on NIST AI RMF and ISO 42001.
ai:governanceAI Red Team Specialist – LLM Red Teaming
LLM red teaming for apps and agents: prompt injection audits, jailbreak tests with garak and PyRIT, OWASP LLM and Agentic Top 10 mapping, defense design.
ai:safety
workflows that use these agents
- Decode: Confusing Document to Plain-Language Action Plan
- Fixit: Home, Car or Pet Problem to Safety-Gated Fix Plan
- Health Brief: Lab Results and Symptoms to Doctor Visit Prep
- Meal Plan: Checked Weekly Plan and Grocery List
- Meeting: Transcript to Traceable Minutes and Action Items
- Postmortem: Incident Evidence to Blameless Report
- Research: Question to Verified Cited Report
- Trip: Feasibility-Checked Day-by-Day Itinerary Planner